Information search apparatus and method

• US 6,526,400 B1
• Filed: 09/30/1999
• Issued: 02/25/2003
• Est. Priority Date: 09/30/1998
• Status: Expired due to Fees

• Current Assignee: Canon Ayutthaya Limited (Canon Inc.)
• Original Assignee: Canon Ayutthaya Limited (Canon Inc.)
• Inventors: Takata, Tomomi, Kobayashi, Yuji
• Primary Examiner(s): Coby, Frantz
• Application Number:
  US09/409,090
• Time in Patent Office: 1,244 Days
• Field of Search: 707/3, 707/2, 707/4, 707/5, 707/10, 707/6
• US Class Current: 1/1
• CPC Class Codes: G06F 16/58 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99936 Pattern matching access

Abstract

An image search apparatus acquires associative words in relation to an input query word, and makes a keyword search of image information on the basis of the obtained associative words and input query word. Parallel to the keyword search, the apparatus acquires a query word and sensory patterns corresponding to the query word, and makes a feature amount search using the feature amounts of the obtained sensory patterns. A search result integration process integrates these search results at weight ratios which are set in advance to obtain final search results. In this manner, the multimedia information wanted can be accurately extracted with respect to the input query word.

Claims

1. An information search apparatus for searching information based on an input query word, comprising:

first search means for determining a query keyword on the basis of the query word, and searching information on the basis of the query keyword;

second search means for determining a feature amount corresponding to the query word, and searching information on the basis of the feature amount;

setting means for setting a search weight to be assigned to search results of said first and second search means; and

integration means for integrating search results obtained by said first and second search means in accordance with the search weight set by said setting means.

2. The apparatus according to claim 1, wherein the search weight includes a first weight corresponding to the search result of said first search means, and a second weight corresponding to the search result of said second search means, and

said integration means applies the first weight to a search matching level of each information as the search result of said first search means and the second weight to a search matching level of each information as the search result of said second search means to obtain an integrated search matching level, and obtains integrated search results on the basis of the integrated search matching level.

3. The apparatus according to claim 2, wherein said integration means selects a predetermined number of pieces of information in descending order of integrated search matching level, and determines the selected information as the integrated search results.

4. The apparatus according to claim 1, wherein said setting means allows a user to set desired weight ratios with respect to the search results of said first and second search means.

5. The apparatus according to claim 1, further comprising:

a weight dictionary which registers weights corresponding to said first and second search means in relation to the query word, and

wherein said setting means sets the weights with reference to said weight dictionary.

6. The apparatus according to claim 5, wherein said first search means derives an associative word associated with the query word, and uses the query word and the derived associative word as query keywords,

said weight dictionary registers weights in units of associative perspectives that connect query words and associative words in units of query words, and

said setting means sets the weights with reference to said weight dictionary on the basis of the query word and an associative perspective designated by a user.

7. The apparatus according to claim 1, wherein the information searched is image data,

said apparatus further comprises:

an image content word holding unit for storing the image data and content words which verbalize concepts expressed in the image data in correspondence with each other; and

an associative word dictionary for storing associative words associated with the content words, and

said first search means acquires an associative word corresponding to the query word from said associative word dictionary, and searches said image content word holding unit on the basis of the acquired associative word.

8. The apparatus according to claim 7, further comprising:

a concept discrimination dictionary for storing index words and antithetic concepts corresponding to the index words in correspondence with each other; and

input means for inputting the query word and a search perspective, and

wherein said first search means acquires an index word and antithetic concept corresponding to the query word from said concept discrimination dictionary on the basis of the query word and search perspective input by said input means, and acquires an associative word corresponding to the query word from said associative word dictionary on the basis of the acquired index word and antithetic concept.

9. The apparatus according to claim 1, further comprising:

a holding unit for storing associative words and sensory patterns in correspondence with each other, and

wherein said second search means acquires a sensory pattern corresponding to an associative word, which corresponds to the query word, from said holding unit, and extracts a feature amount of the acquired sensory pattern as the feature amount corresponding to the query word.

10. The apparatus according to claim 1, wherein multimedia information is image information, and the feature amount is a physical image feature amount obtained by analyzing the image information.

11. The apparatus according to claim 10, wherein the feature amount includes at least one of color scheme information, composition information, and shape information contained of an image.

12. An information search method for searching information based on an input query word, comprising:

a first search step, of determining a query keyword on the basis of the query word, and searching information on the basis of the query keyword;

a second search step, of determining a feature amount corresponding to the query word, and searching information on the basis of the feature amount;

a setting step, of setting a search weight to be assigned to search results in the first and a second search steps; and

an integration step, of integrating search results obtained in the first and a second search steps in accordance with the search weight set in the setting step.

13. The method according to claim 12, wherein the search weight includes a first weight corresponding to the search result in the first search step, and a second weight corresponding to the search result in the second search step, and

the integration step includes the step of applying the first weight to a search matching level of each information as the search result in the first search step and the second weight to a search matching level of each information as the search result in the second search step to obtain an integrated search matching level, and obtaining integrated search results on the basis of the integrated search matching level.

14. The method according to claim 13, wherein the integration step includes the step of selecting a predetermined number of pieces of information in descending order of integrated search matching level, and determining the selected information as the integrated search results.

15. The method according to claim 12, wherein the setting step includes the step of allowing a user to set desired weight ratios with respect to the search results in the first and second search steps.

16. The method according to claim 12, wherein the setting step includes a step of setting the weights with reference to a weight dictionary which registers weights corresponding to the first and second search steps in relation to the query word.

17. The method according to claim 16, wherein the first search step includes a step of deriving an associative word associated with the query word, and using the query word and the derived associative word as query keywords,

the weight dictionary registers weights in units of associative perspectives that connect query words and associative words in units of query words, and

the setting step includes a step of setting the weights with reference to the weight dictionary on the basis of the query word and an associative perspective designated by a user.

18. The method according to claim 12, wherein the information searched is image data, and said method is performed using:

an image content word holding unit for storing the image data and content words which verbalize concepts expressed in the image data in correspondence with each other; and

an associative word dictionary for storing associative words associated with the content words, and

wherein the first search step includes a step of acquiring an associative word corresponding to the query word from the associative word dictionary, and searching the image content word holding unit on the basis of the acquired associative word.

19. The method according to claim 18, wherein said method is performed using a concept discrimination dictionary for storing index words and antithetic concepts corresponding to the index words in correspondence with each other;

and

wherein said method further comprises an input step, of inputting the query word and a search perspective, and

wherein the first search step includes a step of acquiring an index word and antithetic concept corresponding to the query word from the concept discrimination dictionary on the basis of the query word and search perspective input in the input step, and acquiring an associative word corresponding to the query word from the associative word dictionary on the basis of the acquired index word and antithetic concept.

20. The method according to claim 12, wherein said method is performed using a holding unit for storing associative words and sensory patterns in correspondence with each other, and

wherein the second search step includes a step of acquiring a sensory pattern corresponding to an associative word, which corresponds to the query word, from the holding unit, and extracting a feature amount of the acquired sensory pattern as the feature amount corresponding to the query word.

21. The method according to claim 12, wherein multimedia information is image information, and the feature amount is a physical image feature amount obtained by analyzing the image information.

22. The method according to claim 21, wherein the feature amount includes at least one of color scheme information, composition information, and shape information contained of an image.

23. A storage medium for storing a control program which makes a computer search information based on an input query word, said control program comprising:

a code of the first search step of determining a query keyword on the basis of the query word, and searching information on the basis of the query keyword;

a code of the second search step of determining a feature amount corresponding to the query word, and searching information on the basis of the feature amount;

a code of the setting step of setting a search weight to be assigned to search results in the first and second search steps; and

a code of the integration step of integrating search results obtained in the first and second search steps in accordance with the search weight set in the setting step.